浮点数的除法

[SeanQ]

float a = 8.5;
float b = 0.00025f;
unsigned int c = (unsigned int)(a/b);
unsigned int d = (unsigned int)(8.5/0.00025);
执行完以上代码以后(在VC6.0上运行)，为什么c的值为33999,d的值为34000
为什么会出现如此不同的结果

[江南一书生]

unsigned int c = (unsigned int)(a/b);//因为a,b是float类型，除法也是按float（32 bits）精度运算的
unsigned int d = (unsigned int)(8.5/0.00025);//编译器按double类型处理8.5和0.00025，故精度（64 bits）提高
如果用unsigned int e = (unsigned int)(8.5f/0.00025f);则亦指定编译器按float精度处理出发了，则e与c结果等

[江南一书生]

sorry前面打错字了：“按float精度处理出发”应为“按float精度处理除法”